**** BEGIN LOGGING AT Mon Dec 09 02:59:58 2013 Dec 09 04:26:56 Hi, I have BBB, and I'm getting a machine type mismatch from u-boot and kernel. U-boot is 2013.04-rc1-14237-g90639fe-dirty and passing type Dec 09 04:27:08 0x00000E05. Dec 09 04:27:53 Kernel is 3.8.13, and seems to be expecting 0000060a, I guess? Dec 09 04:28:25 Seems funny that this uboot and uImage seem to boot ok from sdcard, but not when I tftpboot the same kernel. Dec 09 04:28:58 I guess that is device tree related? Dec 09 04:49:39 ah, ok. I figured it out. Dec 09 04:50:03 This helped explain it: http://www.informit.com/articles/article.aspx?p=1647051&seqNum=5 Dec 09 05:47:36 how to start programming beaglebone black unsing "C". please guide me to start.. Dec 09 05:48:26 <_av500_> google Dec 09 05:49:15 i have searched a lot. but i didnt know which to use.. Dec 09 05:52:52 i need the correct procedure to access the pins.. help me.. Dec 09 06:16:03 hey guys, do you think you could help me with my bootscript file? Dec 09 06:40:24 is it must to modify the device tree for working with GPIOs in beaglebone black?? Dec 09 06:46:24 anyone help me.. Dec 09 06:55:27 no Dec 09 06:55:37 some GPIOs are already set up for you Dec 09 07:20:49 * somesome1hing cracks sha256 Dec 09 08:48:25 anyone guide me, how to start programing beaglebone black? Dec 09 08:51:20 saravan: the same as you would program on a PC based linux system Dec 09 08:53:30 is there any guide. i'm new to beaglebone.. Dec 09 08:54:32 look for guides on programming on linux Dec 09 08:54:38 not on beaglebone Dec 09 09:30:03 which is the mode number on pinctrl when i want it to be mode 7 and INPUT_PULLDOWN ? Dec 09 09:30:26 0x37 should be input pullup mode 7 Dec 09 12:23:59 hi! Dec 09 12:24:50 I've taken a working NFS rootFS and tried writing it to the eMMC but I can't seem to get a serial console to come up... am i missing something? Dec 09 12:24:53 is it possible to use network sockets from kernel space? need to receive ip multicast messages... Dec 09 12:26:09 Rotti: yes, you can, but you dont want to Dec 09 12:26:25 Rotti: what do you want to do? Dec 09 12:26:47 exosyst: probably wrong kernel boot parameters Dec 09 12:27:17 exosyst: unless you dont even get the bootloader, then it's most probably borked serial port settings, or you overwrote the bootloader too Dec 09 12:27:54 KotH: i'm trying to send a signal to multiple boards. a gpio could be enough but may require more infrastructure... i want this signal to be received at the "same time" Dec 09 12:28:26 Rotti: so you want to control multiple boards using a specific signal? Dec 09 12:28:52 Rotti: what shall happen on the receiving boards? Dec 09 12:29:01 Rotti: who or what triggers the "signal"? Dec 09 12:29:38 that's the question... it could be another board... sending either a multicast packet or just setting a gpio Dec 09 12:32:14 ok. lets go one step back. what is your end goal? what do you want to acheive? Dec 09 12:34:13 KotH: some kind of timesync w/o ntp Dec 09 12:36:10 Rotti: can you tell more Dec 09 12:36:27 i cannot give you hints how to procede w/o knowing what you want to do Dec 09 12:37:59 [13:34] KotH: some kind of timesync w/o ntp Dec 09 12:38:12 KotH: yepp... well, i just want to find out what can be handled/detected with less overhead/faster from the kernel space: a changed gpio or a received network packet Dec 09 12:38:29 well, gpio is faster of course Dec 09 12:40:50 thought so... but i'd like to show it, av500... Dec 09 12:41:56 ? Dec 09 12:43:23 KotH, I'll have a revisit cheers Dec 09 12:45:47 av500: i wrote a small module which 'copies' the value of one gpio to another using an interrupt handler... this time can be measured with a logic analyzer.... for comparison, i'd like to do the same with network packets (i.e. not copy a pin but set the gpio accordingly) Dec 09 12:49:16 Rotti: well.. if you want to sync to a couple of us, then a userspace application using udp packets is enough Dec 09 12:49:22 Rotti: and fearly easy to do Dec 09 12:49:25 fairly* Dec 09 12:49:35 Rotti: if you need better than that, you should go the GPIO way Dec 09 12:50:34 Rotti: with a GPIO pin and an IRQ you get somewhere into the ~100ns range Dec 09 12:50:53 Rotti: with GPIO and capture/compare you get somewhere in the <20ns range Dec 09 12:51:20 Rotti: at this point you want to take care of the rise/fall time of your signal and cable length ;) Dec 09 12:52:52 KotH: gpio and kernel was much better than udp in userspace... but it still was a few us... capture/compare could be interesting ;) Dec 09 12:53:31 Rotti: well, upd packet is mostly dominated by network delay, and that has a jitter of 10-200us Dec 09 12:54:53 Good morning Dec 09 12:55:24 good afternoon Dec 09 12:55:41 KotH: thanks... so i'll just use gpio ;) Dec 09 12:56:25 I'm having trouble getting gpio working. Dec 09 12:56:38 Rotti: what is the use case of your device/sync tool? Dec 09 12:56:52 Actually the gpio is working but I cannot get it to fire a transistor. Dec 09 12:57:36 Using P2222A, by itself, i cannot get output and put two together for a darlington, it dont work either. Dec 09 12:57:49 R2E4_: show us your circuit Dec 09 12:58:46 gpio > 10K emitter to ground collector to load Dec 09 12:59:14 R2E4_: you have a 10k resistor between gpio and base of the 2n2222? Dec 09 12:59:18 gpio > 10K to base emitter to ground collector to loa Dec 09 12:59:27 yes Dec 09 12:59:33 how big is your load? Dec 09 12:59:47 just a small relay Dec 09 12:59:53 how many mA? Dec 09 13:00:07 100ma Dec 09 13:00:12 well... Dec 09 13:00:20 your base resistor is too high Dec 09 13:00:33 the 2n2222 has a beta of 100 if i'm not mistaken Dec 09 13:00:44 ie, you need a base current of at least 1mA Dec 09 13:00:53 you cannot get that current with 10k and 3V Dec 09 13:02:14 also make sure you havent forgotten the free-wheel diode for the relay, otherwise you'll fry the 2n2222 ;) Dec 09 13:02:18 KotH: no real use case yet... i want to see how accurate clocks can be synced in a distributed system and if it is possible to use something like this for some delay measurements etc... Dec 09 13:02:45 Rotti: 1ns is "easy" to acheive Dec 09 13:02:52 300 ohms? Dec 09 13:03:34 Rotti: but, when you go below 100ns you have to take care of a lot of things, otherwise the disturbances will be bigger than what the boards can do Dec 09 13:03:53 R2E4_: <3k Dec 09 13:04:02 R2E4_: i'd use 1k, then you have a factor of 3 Dec 09 13:04:09 R2E4_: but check the beta of the 2n2222 first Dec 09 13:04:30 its 50ma Dec 09 13:04:41 oh no, 1 Dec 09 13:04:45 KotH: i guess ~anything <1us is good enough for now... Dec 09 13:04:47 beta is just a number, not a current ;) Dec 09 13:05:38 R2E4_: beta is often labled as h_fe in datasheets Dec 09 13:05:58 http://www.fairchildsemi.com/ds/PN/PN2222.pdf no ref to it Dec 09 13:06:37 300 max Dec 09 13:07:37 R2E4_: get yourself a copy of http://www.amazon.de/Halbleiter-Schaltungstechnik-Ulrich-Tietze/dp/3642310257/r Dec 09 13:07:40 R2E4_: get yourself a copy of http://www.amazon.de/Halbleiter-Schaltungstechnik-Ulrich-Tietze/dp/3642310257/ Dec 09 13:08:08 Do I need to learn german first? Dec 09 13:08:11 yes Dec 09 13:09:38 well, you can also buy the english edition, but it's based on a 10y old german edition Dec 09 13:10:09 * das remembers when he saw the book for 30 euros and didn't buy it ... Dec 09 13:10:14 although the transistor circuits didnt change much, the sections were heavily expanded in the newer german editions Dec 09 13:10:30 I just want to send ground through the NPN and directly to a relay, and 12v on the other side of relay. THis is not rocket science. Dec 09 13:10:40 das: the german edition cost not even half the price.. for a new book Dec 09 13:10:53 das: used ones are usually 30 euro or so Dec 09 13:11:11 R2E4_: yes, it's not Dec 09 13:11:12 KotH: T&S ? last time I checked it was >100E new ? Dec 09 13:11:19 R2E4_: but there are still 100s of ways to do it wrong Dec 09 13:11:20 :) Dec 09 13:11:37 R2E4_: as i said, decrease the base resistor, add a free wheel diode to the relay and you are done Dec 09 13:11:59 das: i got it for 120CHF or so.. about a year ago Dec 09 13:12:31 dat 1.20 rule :) Dec 09 13:12:37 R2E4_: oh.. and make sure the 2n2222 gets enough cooling. 100mA is a considerable current Dec 09 13:12:59 das: chf, the only stable currency in europe ;) Dec 09 13:14:12 well it's "indexed" on the euro Dec 09 13:14:30 because it was too stable ^^' Dec 09 13:15:03 that's the reason ? Dec 09 13:15:10 hum I'll read up on that Dec 09 13:15:37 das: yes, chf skyrocketed relative to eur, which had a huge imapact on the largely export based economy of ch Dec 09 13:15:57 das: thus they had to devaluate chf in order for the swiss economy to remain stable Dec 09 13:16:53 I fail to see how it makes the chf stable Dec 09 13:17:22 people were buying chf because it was more stable than eur Dec 09 13:17:33 that increased the price of chf relative to eur Dec 09 13:18:11 well by buying a currency you increase the pool of money in this currency thus increasing its value Dec 09 13:18:28 so that means that people trusted the chf so much that it became unstable Dec 09 13:18:33 relative to other currencies Dec 09 13:18:47 since I doubt most chf were bought in eur Dec 09 13:19:14 for that, you have to ask someone who knows more than i do about forex Dec 09 13:19:22 drpooed the resistor to 1K and it still not working. Dec 09 13:19:56 R2E4_: ok.. what do you measure? Dec 09 13:20:08 KotH: I think we (I?) confuse stability of the currency itself and stability of the state that backs up the money :) Dec 09 13:20:21 das: heh.. could be :) Dec 09 13:20:30 bitcoins ftw Dec 09 13:20:47 Crofton|work: i thought you use usrp's as currency? :) Dec 09 13:20:54 :) Dec 09 13:26:18 IT is showing 1.2volts instead of 12vdc. Not enough current? Dec 09 13:33:45 R2E4_: IT? Dec 09 13:33:49 R2E4_: what are you measuring? Dec 09 13:38:58 ground on the collector and 5 volts Dec 09 13:43:27 * KotH smiles and nods Dec 09 14:37:58 hi, i am trying to boot a BBB from the micro-SD card and all I get is the letter C printed a few times a second when I view the boot process using a UART. I verified the write of the OS on the SD card and it looks alright. how do I fix this ? Dec 09 14:46:38 fix the card Dec 09 14:46:51 also, hold the boot button at power up Dec 09 15:22:58 av500: when you say fix the card you mean re-install the OS on the card or use a new card ? Dec 09 15:23:22 av500: i tried pressing the Boot button and then restarting the BBB but that didn't change anything. Dec 09 15:26:45 it will boot from a properly prepared card Dec 09 15:26:51 it does so for many people Dec 09 15:29:51 If the BeagleBone is powered via the 5V Power supply and the computer is disconnected will the USB Mini Receptacle have power? Dec 09 15:30:07 The USB Mini Receptacle located on BeagleBone Dec 09 15:32:12 no Dec 09 15:32:19 there are diodes Dec 09 15:46:41 has anybody some experience creating a BBB web server with a simple web interface? PyBBIO (https://github.com/alexanderhiam/PyBBIO) seems quite straightforward, suggestions? comments? Dec 09 15:51:48 vmayoral Dec 09 15:51:56 did you ask a question about the web server? Dec 09 15:52:00 yes Dec 09 15:53:05 I used python as a webserver Dec 09 15:53:05 for one of my projects Dec 09 15:53:16 did you use the PyBBIO? Dec 09 15:53:20 library* Dec 09 15:53:40 Not for this project Dec 09 15:53:56 http://docs.python.org/2/library/simplehttpserver.html Dec 09 15:54:25 i see that's also another way, any recommendations for starting a new project? Dec 09 15:54:59 i mean, since you've used both Dec 09 15:55:10 I use both Dec 09 15:55:19 PyBBIO mainly for simple GPIO stuff Dec 09 15:55:45 In my application i honestly don't need much, i get the data from UART and I2C Dec 09 15:56:00 so i just want to display it and let others display it with their devices Dec 09 15:56:41 Since PyBBIO does not have a webserver built in, it would be hard to use for that purpose. Dec 09 15:57:33 If it were up to me, I would use Python webserver just because it is very basic; and if you need it to be complex you can Dec 09 15:58:32 ok, i thought PyBBIO had a web server built in Dec 09 15:58:36 my mistake then Dec 09 15:58:40 I've never used PyBBIO webserver Dec 09 15:58:47 I think it has one. I just never used it Dec 09 15:58:50 thanks for the feedback :) Dec 09 15:59:01 i think i'll give a try your way bpetersen Dec 09 15:59:18 thanks to both! Dec 09 15:59:22 awesome! Dec 09 16:00:09 Balls, it does. Dec 09 16:00:24 Why the hell does a GPIO library have a webserver in it? Dec 09 16:01:54 restfulness ? Dec 09 16:02:18 I have no idea what you think that means. Dec 09 16:02:33 web2.0 all the things? Dec 09 16:03:30 BBIOServer Dec 09 16:03:31 separate threading, separate app, able to keep it live and active and interface via REST Dec 09 16:03:33 I am just saying, there are perfectly servicable Python web server modules. Why add one to a module that should have nothing to do with that? Dec 09 16:04:10 you should be able to understand the usefulness of that if you have much experience Dec 09 16:04:26 This would be like, "Hey, here is a blender. It blends stuff. But, if you want, it is also a motorcycle." Dec 09 16:04:45 that makes no sense Dec 09 16:05:00 thews: Right, but there is no reason for the *module* to have the built in. It is the programmer's job to use the modules s/he needs to make the application s/he wants. Dec 09 16:05:18 agmlego: if it can be enabled/disabled it's quite useful Dec 09 16:05:23 don't have to reprogram it yourself Dec 09 16:05:36 also it allows test cases to be built and give people examples Dec 09 16:05:45 When I want to use gpio, I use a gpio module. When I want to add a web server to the rpogram, I use a web server module in addition to the gpio module. Having them the same module is just messy and poor practice. Dec 09 16:05:46 and make sure it works properly with each build Dec 09 16:06:12 But then again, this is the open sores community. :-/ Dec 09 16:06:31 sure. Dec 09 16:06:52 Poor practices are unfortunately part of the territory. Dec 09 16:06:55 something tells me you don't understand test frameworks either Dec 09 16:07:14 thews: Something tells me you do not understand the complaint I am making. Dec 09 16:07:42 agmlego: I understand it, but it shows lack of experience Dec 09 16:08:14 thews: ...yeah, no it does not. Dec 09 16:08:25 sure. Dec 09 16:08:37 One purpose per module, one module per purpose. Dec 09 16:08:44 THis is standard programming practice. Dec 09 16:09:11 and what proprietary library do you use that adheres to this? Dec 09 16:09:12 It is all well and good to have a framework of multiple modules. Last I checked, though, PyBBIO is a single module. Dec 09 16:09:31 thews: Every one I have ever touched, or written, excpet apparently PyBBIO. Dec 09 16:10:05 who uses your "modules" ? Dec 09 16:10:25 Everyone on my team. Dec 09 16:10:37 I'm sure you think this is useless too http://electricimp.com/ Dec 09 16:10:48 Also, if I may ask, how much experience with Python do you have, thews? Dec 09 16:11:01 thews: I have no opinion on the Imp. Dec 09 16:11:24 python? about 10-12 years Dec 09 16:11:40 easy to deploy gpioasaservice.com Dec 09 16:11:53 mdp: absolutely Dec 09 16:11:59 same as the electric imp Dec 09 16:12:03 it's quite valuable Dec 09 16:12:06 mdp: Just as easy to bolt on a dedicated web server module and do it the right way. Dec 09 16:12:06 everything is justifiable in python-land Dec 09 16:12:13 No it really is not. Dec 09 16:12:21 agmlego: not just as easy Dec 09 16:12:28 so make a sane module Dec 09 16:12:30 Sure it is. Dec 09 16:12:31 heh Dec 09 16:12:33 "you have the source" Dec 09 16:12:37 fuck it and go Dec 09 16:12:42 mdp: Or I can bitch about it in IRC. Dec 09 16:12:42 he doesn't like open sores Dec 09 16:12:49 I thought that was what this channel was for. ;-P Dec 09 16:13:02 he will make his own and not use the experience and evolution of others Dec 09 16:13:10 this channel is for vague unanswerable questions that are never asked Dec 09 16:14:10 thews: Who said that? My dig at the open source community is with the frequent value judgement prioritising speed and release over quality or planning. Dec 09 16:14:34 agmlego: sounds like you utilize bad opensource projects Dec 09 16:14:45 ...like PyBBIO, yeah. Dec 09 16:14:46 ;-P Dec 09 16:15:20 agmlego: I often find the opposite. I think it depends what scale of project you're dealing with Dec 09 16:15:25 many have timelines, features, test cases to guarantee reliability, many use code coverage, many use best practices checkers, all built in, some use on the fly testing like Guard Dec 09 16:15:46 agmlego, it does what the author wants..deal with it Dec 09 16:16:12 thews: Many do, many also do not. Dec 09 16:16:24 mdp: I am, by bitching about it in IRC. Dec 09 16:16:40 mdp: THis is how people on the Internet deal with things. Deal with it. Dec 09 16:16:41 ;-P Dec 09 16:17:05 agmlego, be sure to post a rant on g+ too Dec 09 16:17:55 Yeah, I probably ought to do that. Dec 09 16:19:01 if the author is on there and you include an ad hominem attack that should go a long way to addressing the issue Dec 09 16:19:23 Yeah, and I can probably build a strawman for the purpose too. Dec 09 16:19:36 I think I still have a partial bale out back. Dec 09 16:19:42 or...just switch to libsoc Dec 09 16:19:52 Heh. Dec 09 16:20:55 the advice is all free and you get what you pay for Dec 09 16:20:56 And SKOLD vs. KMFDM - Bloodsport comes up in my random play list. How appropriate. Dec 09 16:21:10 Honestly though, I may well be writing my own Python gpio module soon anyway, because as far as I have found, all of the Python ones only talk through sysfs which is too slow for much of what i want to do. Dec 09 16:21:29 agmlego, hackaday has a perfect post for you then ;) Dec 09 16:21:35 I think you mean writing a kernel module then? Dec 09 16:21:47 mdp: I literally loaded that page about three minutes ago. Dec 09 16:22:10 jackmitchell: A kernel module with a Python userspace wrapper, yes. Dec 09 16:22:14 jackmitchell, no way, he wants to open /dev/mem and bang away ;) Dec 09 16:22:15 kernel gpio subsystem is fast, sysfs isn't too slow though Dec 09 16:22:28 thews: kHz is slow. Dec 09 16:22:35 I need MHz. Dec 09 16:23:07 agmlego: if you need MHz you probably want to write it out in C anyhow Dec 09 16:23:25 agmlego, take it to 11 by using the PRU Dec 09 16:23:47 thews: Why would I want to do that? Dec 09 16:23:54 mdp: Also working on that too. Dec 09 16:23:59 Or will be., Dec 09 16:24:46 agmlego: you'll find the BBB cpu to be underwhelming MHz operations on a scripted language on 1000MHz CPU can be interesting Dec 09 16:25:18 and 1000MHz CPU that lacks lots of good instruction sets Dec 09 16:25:23 thews: I think that it will be fine, actually. Dec 09 16:25:34 AVX FPU would be nice Dec 09 16:25:53 I just need to be a touch closer to the hardware than the sysfs abstraction layer. Dec 09 16:27:57 thews: Also, I have a fair amount of experience with the CPU on the bones. It is a perfectly servicable device. Dec 09 16:28:07 *CPUs Dec 09 16:29:40 agmlego: I have a little bit of experience with different instruction sets and limitations , not nearly as much as some kernel developers and scientific devs though Dec 09 16:30:14 thews: The good news is, I am not aiming at doing high-performance computing on a $40 SBC. Dec 09 16:30:26 I am using it as a realtime controller, which is what it was designed for. Dec 09 16:31:16 realtime has a lot of meanings these days Dec 09 16:31:39 thews: I am aware. Dec 09 16:31:49 I get paid to do industrial automation. Dec 09 16:35:23 agmlego: what is the latency of python responding to real-time interrupts? Dec 09 16:36:00 Have not measured that yet. Dec 09 16:36:33 agmlego: what about IO blocking times? Dec 09 16:38:32 thews: It is difficult for me to tell whether you are genuinely curious about a subject I have stated that I have not yet done the research on to have solid figures for, or if you are just being deliberately combative to try and prove some unknown point. Dec 09 16:38:55 TL;DR: Not sure if real or troll. Dec 09 16:39:01 agmlego: I am genuinely interested, to be helpful to you, I recommend you check out stackless Dec 09 16:39:22 but I don't think you'll possibly get any real-time usefulness out of python, because of the issues above Dec 09 16:39:33 sysfs will be less limiting than python itself Dec 09 16:39:43 python thread creation for one is terribly slow Dec 09 16:40:06 I think you are underestimating the interpreter, and also naively assuming that I will be using the default one with no performance modifications. Dec 09 16:40:14 But thank you for your unrequested advice. ;-P Dec 09 16:40:20 agmlego: I just have experience in these things Dec 09 16:40:26 Yes. As do I. Dec 09 16:40:38 A big part of the automation I get paid to do, is done with Python. Dec 09 16:40:40 agmlego: I've worked with timing constraint stuff for emergency systems for the last 8 years Dec 09 16:41:03 ANd I have worked with high-speed machine vision for the last 6. Dec 09 16:41:25 another place where python would not be well suited, maybe if you used C extensions Dec 09 16:41:25 So lets both put our penises back in our trousers, and accept that each other may actually have a clue about this stuff. ;-P Dec 09 16:42:26 Python works just fine for all the glue work around the system, including acting as the communications and control interfaces. Dec 09 16:42:53 The actual image-processing is not done with Python, but basically everything else is. Dec 09 16:44:04 I do similar things, but I call it scripting, because I feel like that I'm being honest with myself in those cases and I prefer the ease of programming with scripted languages, and the code clarity, and rapid development Dec 09 16:44:51 A rose by any other name, but if that helps you sleep at night, whatever. Dec 09 16:44:56 but I can't handle 1M iops in scripted languages Dec 09 16:45:02 I'm honest with myself there Dec 09 16:45:17 Right, you said that already. Dec 09 16:45:28 IF you want to do real time programing in python, GO HOME. Dec 09 16:45:35 But you say it in a way that makes it sound like you are accusing me of somehow lying to *myself*. Dec 09 16:45:50 agmlego: if that's what you feel Dec 09 16:45:56 But then again, I am arguing with someone on the Internet. Dec 09 17:09:57 This is a slightly random question but how long should I test for over temperature protection of a device? Would 4 hours be sufficient or should I go for like 24 hours? Dec 09 17:11:25 bpetersen: what are you testing? Dec 09 17:13:45 bpetersen: and for what reason? I imagine for QA testing, you'd only need to test it once in quantity, is this for TCO? Dec 09 17:14:31 bbiab Dec 09 17:18:25 its to test a motor Dec 09 17:18:54 small one about 1 foot tall Dec 09 17:19:24 Yea QA testing Dec 09 17:22:23 Are there more IRCs for electrical engineering? Dec 09 17:46:42 bpetersen: not sure for thermal stuff you'd need to give it quite a load, and over drive it Dec 09 17:47:23 bpetersen: lots of stuff in the channel list with electr* Dec 09 18:00:51 "small [motor]"..."about *1 foot tall*" Dec 09 18:01:04 Uh, yeah, in this channel, that would be a *massive* motor. Dec 09 18:12:13 does anybody know when the beaglebone is back in stock and how many will be available? Is this a hard to get item? Dec 09 18:12:59 more are shipping to distributors all the time and CCo is splitting up shipments to all distributors end up getting some. Best to place an order early and be next in line. Dec 09 18:13:26 I see Dec 09 18:20:06 ANd if you really do not wqant to wait, basically all the componets are in-stock at Digikey, and you can get the board made at your local PCB fab house of choice. ;-P Dec 09 18:21:50 trav: basically all the stock was purchased up to supply to microcenter and radioshack recently for the christmas shopping season Dec 09 18:21:59 trav: we are ramping up production to compensate Dec 09 18:22:18 trav: you should see inventories updated within the next few days at most of the distributors Dec 09 18:30:31 jkridner: ping Dec 09 18:30:36 ??? Dec 09 18:33:09 prpplague: ???? Dec 09 18:40:34 jkridner: did you hear back from TI on the "customization" ? Dec 09 18:58:03 thews: customization? Dec 09 18:58:22 sgx driver Dec 09 19:02:51 thews: ahh Dec 09 19:38:11 Hi everyone, I was interested in some help with a basic cross-compilation error for Beaglebone Black. This is in Eclipse, in Ubuntu, I'm getting a Symbol 'cout' could not be resolved error. I have all of my include and library path directories set up correctly, and I can see a list of functions when I type "std::" but cout is not there ?! I'm new to eclipse and linux programming, my Libraries field is empty, could this be the caus Dec 09 19:44:29 is included, c++ versions are 4.6 and 4.6.3 Dec 09 19:48:16 brb,, Dec 09 20:38:24 hmm still looking for help.. thanks, Dec 09 20:40:06 you included iostream.h? Dec 09 20:40:13 mckennetho Have you seen this tutorial from Derek Molloy? http://derekmolloy.ie/beaglebone/setting-up-eclipse-on-the-beaglebone-for-c-development/ Dec 09 20:41:15 this is the one I am following Dec 09 20:42:25 I am trying to configure Eclipse in windows (without a linux vitual box) for cross development. Close but no cigar yet. Dec 09 20:42:28 set up exactly the way it is in the tutorial, yes iostream is included Dec 09 20:43:11 hmm, did you set up c++ manually, or use build essential? Dec 09 20:44:22 I just did the Hellow World C++ project, if that's what you mean, but I had to include all of c++ directories, in the video they were already included. Is this an include or a library error? Dec 09 20:45:14 Int he video you can't see what libraries he has included :/ Dec 09 20:46:15 can you pastebin your cpp file? Dec 09 20:49:01 It's just 3 lines Dec 09 20:49:23 #include Dec 09 20:49:28 using namespace std; Dec 09 20:50:00 int main(){ cout<< "test" < hi Dec 09 20:50:38 you'll need a space between the return and the 0 Dec 09 20:51:10 one question : can i use gpio in android ? Dec 09 20:52:53 <_av500_> sure Dec 09 20:54:41 any link ? Dec 09 20:55:34 sorry that was a typo, I'm on a pc, compiling on a different machine Dec 09 20:55:51 <_av500_> Android is Linux Dec 09 20:55:57 <_av500_> linux can use GPIOs Dec 09 20:56:31 <_av500_> but here: https://www.google.com/search?q=android+gpio Dec 09 20:56:48 do you know if this is a c or c++ include error? And what is build essential? Dec 09 21:02:56 _av500_, which link will work? Dec 09 21:09:05 If I build a regular project that is not cross-compiled, everything works well Dec 09 21:13:39 the error seems to happen when I include /usr/arm-linux-gnueabi/include/c++/4.6.3 under c++ includes Dec 09 21:15:30 did you use the cross settings for your project? Dec 09 21:18:05 does anyone know how to programatically cut power to the host port on a BBB? I have a device that I'm interfacing with via libusb that sometimes misbehaves and needs to be physically disconnected/reconnected in order to function again. Dec 09 21:18:26 <_av500_> not sure you cut power Dec 09 21:18:33 <_av500_> did you check the schematics? Dec 09 21:20:53 yes, hmm I just noticed that when I do a build, eventhough it gives me those "symbol could not be resolved" fatal errors, it still compiles an executable Dec 09 21:21:56 and I just tested the executable, and it worked just fine, so this has to be an Eclipse incompatilbity bug. It's the newest version too. Dec 09 21:25:57 sorry nm, even though it compiles when it shows error, it compiles a previous version and it doesn't show my changes, I don't understand this IDE Dec 09 21:34:38 eclipse works fine for cross compiling, if you have a cross toolchain configured and make the project use it Dec 09 21:36:30 I have, I figured it out but I still can't explain why it shows fatal errors but compiles and works :/ must be a bug in eclipse Dec 09 21:37:10 if you cancel out of a debug build it does that occasionally, restarting Eclipse resets it Dec 09 21:45:31 I'm running Angstrom Linux beaglebone 3.8.13 #1 SMP. I'm attempt to work this bonescript example for a shift register http://beagleboard.org/support/BoneScript/shiftOut/ Dec 09 21:45:55 Not sure if my bonescript supports it. Dec 09 21:46:12 beaglebone:~# opkg list-installed | grep bonescript Dec 09 21:46:12 bonescript - 1.0-r21.5 Dec 09 21:46:12 beaglebone:~# Dec 09 21:46:15 beaglebone:~# opkg list-installed | grep bonescript Dec 09 21:46:16 bonescript - 1.0-r21.5 Dec 09 21:46:16 beaglebone:~# Dec 09 21:46:53 I note this in the tutorial: Note: This doesn't work until BoneScript version 0.2.3 Dec 09 21:47:18 GumShoe: please don't paste more than one line here Dec 09 21:47:30 GumShoe: if you need more lines, please use a service like pastebin Dec 09 21:48:35 looks like I'm running 0.2.15 ??? Dec 09 22:21:58 I'm losing it, or this diagram is wrong. About 60% down the page an image of the BBB with P8 and P9 reversed/mislabeled http://letsmakerobots.com/node/37063 Dec 09 22:23:10 why dont you use the official pdf '? Dec 09 22:23:33 I'm looking through that now. Dec 09 22:27:13 GumShoe, there are conflicting sets of pinouts. refer to the manufactures page Dec 09 22:42:43 Has anyone had any luck getting CentOS on a beagleboard-xm? Dec 10 00:08:31 What is a cape? Is it basically a sensor for the board? Dec 10 00:14:43 Cape:Bone::shield:arduino Dec 10 00:20:04 agmlego: Too slow :) Dec 10 00:26:07 Hello :D Dec 10 00:28:14 Hy bordicon Dec 10 00:28:35 Hey BellinganRoy Dec 10 00:29:30 anyone have experience with the craneboard? I am evaluating it for possible deployment on a cubesat Dec 10 00:30:05 Running into issues getting serial console output on uart3 Dec 10 00:46:10 Hi everyone, I have a question regarding opening the led brightness file on the beaglebone. If I use the command ledHangle = fopen("/sys/class/leds/beaglebone:green:usr3/brightness/") it returns a NULL to the LED handle. I know the file is there, is this a permissions issue? Dec 10 00:56:18 also does anybody know how to stop gdb-multiarch ?? Dec 10 00:56:44 if it is listening on a port and I want it to stop because I can't start the degubber on the client Dec 10 00:57:47 Anyone successfully configured an IDE on windows for cross development to the BBB? Looking for open as in free toolchains. I've been looking at Linaro but not clear to me if it will work. Dec 10 00:57:51 https://launchpad.net/linaro-toolchain-binaries/+milestone/2013.11 **** ENDING LOGGING AT Tue Dec 10 02:59:58 2013